Although Institutional review boards (IRB) were originally intended to reflect local values and knowledge; they have been widely criticized as inappropriate for large multi-site studies. One proposed solution is increased use of central IRBs (C-IRBs), which would manage human subject protection for all of the sites. This study will use innovative methods and built on prior research on IRBs by the investigators to describe how C-IRBs review multi-site projects, and how these reviews differ from single site reviews and the reviews of multi-site projects by local IRBs. The results of this study are intended to help (1) define the principles and practices involved in the conduct of CIRBs, (2) understand whether different tools or procedures will help to maximize the effectiveness of CIRBs and (3) describe any ethical and logistical issues that should be taken into account when IRBs act as C-IRBs. This proposal combines methods that we used in two different studies of local IRBs to describe central IRBs decision-making processes. These methods will include audio-recording of and estimated 249 reviews at 22 different IRB panels associated with 10 different central IRBs and semi-structured interviews with administrators, IRB chairs, reviewers, and staff members. We have developed systematic procedures for coding and processing this type of data so that we can do both systematic qualitative and sophisticated quantitative, analyses of the data. It will allow us to describe the organization and decision making of central IRBs. In addition, we will explore approaches for determining how central IRBs compare to local IRBs in the topics considered during the reviews of protocols requiring full board review. Our research on local IRBs has made us uniquely qualified and ready to study this innovation IRB review.